Research on Optimization of the Distribution System Consisting of Multiple Manufacturers and One Retailer Based on Q-Learning

摘要

This paper studied a distribution system consisting of multiple manufactures and a single retailer. In this system, in order to motivate the retailer, the manufacturers must decide how much bonus will be provided. The paper analyses the factors of profit formation of manufactures and retailer. This system was simulated based on Q-learning algorithm. The results show the manufactures’ profits are damaged in non-cooperation situation. The manufactures would be benefited as they adopt the cooperation action, but the retailer’s profit is damaged.